Health facility data for monitoring of maternal, newborn and health indicators: a systematic approach applied by 32 countries in sub-Saharan Africa

Background Routine health information system (RHIS) data in sub-Saharan Africa are a critical but underused source for maternal, newborn, and child health (MNCH) monitoring. Rapid digitization through DHIS2 creates an opportunity to standardize data extraction, data-quality assessment, adjustment, and analysis. In this study, we demonstrate how the Countdown to 2030 for Women’s, Children’s and Adolescents’ Health (CD2030) collaboration has fostered cross-country partnerships across 32 sub-Saharan Africa and developed a harmonized data extraction and analysis platform to enhance the systematic use of RHIS data. Methods CD2030 developed an integrated data suite: (1) a standardized DHIS2 extractor that retrieves and maps MNCH indicators into a harmonized template and (2) an analytical application that assesses data quality at national, regional, and district levels, including adjustments for reporting completeness and outliers. The suite was implemented mainly as an R Shiny application and deployed in 2025 with country teams from Ministries of Health and public health institutions in 32 sub-Saharan African countries. Results Despite variation in DHIS2 metadata and indicator definitions, all 32 country teams used the application; 30 extracted 2020–2024 data successfully, and all applied analytical modules. Median overall data-quality scores declined from 83.7% (2020) to 79.3% (2024), with low 2024 scores in Chad, Central African Republic, South Sudan, Guinea-Bissau, Ghana, Senegal, and Somalia. Reporting completeness remained high throughout (median ≥ 95%), though Senegal and South Sudan were lower in 2024. Inconsistencies persisted between ANC-1 visit counts and first-dose DTP counts when compared with ratios expected from household survey estimates. Extreme outliers were rare (mainly Guinea and Somalia) and adjustments had negligible impact on estimates. Facility-derived denominators generally outperformed population projections: Penta-1–derived denominators were most often selected for immunization coverage, while institutional delivery typically relied on ANC-1–derived denominators. For Penta-3, 16/22 countries were within ± 5 percentage points of survey estimates (median absolute difference [MAD] 3; IQR 1–5). For institutional delivery, 15/21 were within ± 5 points (MAD 4; IQR 2–7). Conclusions CD2030’s harmonized DHIS2 extraction and analytics enable more standardized use of RHIS data for MNCH monitoring. Combining routine data with household surveys provides complementary, more frequent, and more localized insights to support health system performance monitoring.